Performance analysis of roof-mounted photovoltaic systems on the office building in HanoiSubmitted on 7 Dec 2016
HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
Copyright
Performance analysis of roof-mounted photovoltaic systems on the office building in Hanoi
Xuan Truong Nguyen, Quang Hung Nguyen, Dinh Quang Nguyen
To cite this version: Xuan Truong Nguyen, Quang Hung Nguyen, Dinh Quang Nguyen. Performance analysis of roof- mounted photovoltaic systems on the office building in Hanoi. 4th Regional Conference on Energy Engineering 2016, Institute of Technology of Cambodia (ITC), AUN/SEED-Net Regional Conference on Energy Engineering, Nov 2016, Phonm Penh Cambodia. pp.41-46. hal-01394259v2
41
systems on the office building in Hanoi
Nguyen Xuan Truong*, Nguyen Quang Hung, Nguyen Dinh Quang
Clean Energy and Sustainable Development Laboratory, Energy Department
University of Science and Technology of Hanoi, VAST
Hanoi, Vietnam
Abstract— Building is responsible for more than 40 percent of
energy consumption and one third of greenhouse gas emissions.
Europe has established the pathway towards nearly zero-energy
buildings (nZEB), soon required in every new construction and
large renovation in existing buildings by 2020. Achieving nZEB
required the use of photovoltaic modules for generating
electricity. The rooftops of residential, commercial and office
building are ideal places for the installation of PV systems.
Recently, in Vietnam, the nZEB becomes an important part of
sustainable development and attracts the attention of scientists
on the term of “green building” or “smart building”, while
photovoltaic systems are in most cases not used in building today.
From this framework, we aim to implement the platform that
will be dedicated to prototyping and tools to support research
and training in the reference building. The first work presented
in this paper aims at parametric analysis of 15 kWp PV system
applied to a rooftop of a University building in Cau Giay - Hanoi.
The study shows the effect of PV module types, and installation
tilt angle on the electrical energy yield. The Photovoltaic
simulation tool takes an important role in calculating the energy
production from the solar panels. PVSyst software as a
simulation tool was used to simulate several operating conditions
as well as components. The research investigated a complete on-
grid PV system by taking into account orientation of PV array
under meteorological condition in Hanoi, solar cell technologies
and inverter connections. Meteorological data was accessed by
MeteoNorm and PVGIS. According to simulation results, the
configurations of PV system for reference office building of the
University of Science and Technology of Hanoi (USTH), was
optimized to facilitate the next phase of smart building research in Vietnam.
Index Terms— green building; photovoltaic source; simulation
tool; rooftop; energy efficiency.
concluding that this sector represents more 40 % of total
energy use in EU and USA. Most of this energy is for the
provision of lighting, heating, cooling, and air conditioning.
The green and sustainable development movement in the world
has placed green building in high priority as it is able to meet the building demand while mitigating the negative impacts of
construction sector. High energy performance level in building
can be attained by taking advantage of passive design
techniques and active solar technologies such as solar
collectors for hot water and space heating, PV panels for
generating electricity. PV array installed in the building
contributes to the reduction of the primary, conventional
energy supply, as well as the reduction of CO2 emissions in the
surrounding environment. Distributed PV system can be
installed as façade or rooftop applications in order to maximize
the benefits of clean and quiet power generation. Façade
installation can be optimized by observing the orientation and
distance to length ratio. Rooftop application can observe the
tilted angle and curved installation [2, 3].
Energy consumption in Vietnam’s building sector is
growing quickly as a result of rapid industrialization and
urbanization. In the report [4] have shown that, the residential sector is one of the main energy consumers in Vietnam,
accounting for 33% share in 2012 and 35.6% in 2014, which
will continue to increase in the future due to the pressure of
increasing demand for buildings. The nZEB has become an
important part of sustainable development and attracts the
attention of scientists on the term of “green building” in
Vietnam. However, the adoption of green building is still
limited, the release of related regulations are slow and
governmental support are lacking. It is recommended that the
Vietnamese government should take stronger actions such as
ratifying regulations or offering incentives to promote green building for sustainable development [5-7]. Besides, it has
limited programs addressing renewable energy like PV
generation deployment and energy efficiency use. Generally,
the photovoltaic systems are yet to be used in the form of
building-integrated systems nowadays. Such system is still at
the research and demonstration stage. In this regard, it is
required that the design and control strategies of green building
are not straight forward since the buildings may involve
complex integration of renewable sources, energy appliances,
energy storages and may interact with the local smart grid.
Supported by the host university, our research project is a
contribution for a better understanding of the problem. This
paper mainly aims to: (1) present a platform for green building
research and training; (2) estimate the building energy demand for lighting, HVAC, elevator, and other equipements; (3) use
the PV system simulation software PVSyst to simulate
performance of the 15 kWp PV system and investigate monthly
4th AUN/SEED-Net Regional Conference on Energy Engineering 2016 DRE02
42
meteorological data of the installation site and PV array.
Finally, simulation results will be used to advise the
configurations of PV system for the building of the host
university to contribute for the next phase of green building
and nZEB research in our laboratory.
II. PLATFORM FOR GREEN BUILDING RESEARCH
AND TRAINING AT USTH
A smart micro-grid platform at local level (building) with integrated Photovoltaic array and building energy management
system is necessary to support green building research and
eventually improve the current power grid by augmenting its
performance and energy efficiency. This platform allows to
study many aspects of smart building in Vietnam including
building modeling with analysis of the differences between the
model and the real system, understanding of distribution of
energy in management issues under different scenarios. The
power supply for this platform consists of a 15kWp PV
generator to be installed on the rooftop, an electrical storage
system and the power grid. The system extracts the maximum
power obtainable from the PV array under different working conditions to provide a portion of the building power demand.
The loads include the lighting system, HVAC system,
laboratory equipment and the elevator.
Besides, the monitoring system allows data to be collected
and utilized for the optimal operation. The energy controller
reads all the data measured by the transducers to operate the
equipment according to different selected modes. The energy
manager controls the delivery of energy, and the charge or
discharge of batteries.
The objective of the project is to benefit from PV power
supply, to improve energy efficiency in the building and to
reduce the cost of electrical consumption under different
scenarios.
III. DATABASE COLLECTION
To set up a reliable PV system, it is important to identify the important data which affect the performance of the system.
The data collection for this study divided into two categories
which are the meteorological database and solar radiation, and
total energy used by requirement loads.
A. Meteorological database and solar radiation
The building is located at N°.18, Hoang Quoc Viet Street,
Cau Giay District, Hanoi, in the northern of Vietnam at a
latitude of 21°N. It has one 37.9 meter high block, with a
939 m2 rooftop consisting of three available installation areas. Total surface area for two flat roof is 420 m2 and a flat (free
standing system) with an available area of 130m2, where PV
panels can be installed (rooftop-mounted PV system).
The solar radiation climate database of a zone is extremely
important to estimate the performance of a PV generation
system. Vietnam is located in South East Asia, extending
between latitudes 9°N and 23°N. The geographical location of
the building and the local weather conditions influence the
optimal tilt of the PV modules and is, therefore, of great
importance. Several studies that addressed the solar resource
maps show that global horizontal irradiation in annual daily average reaches around 3.4 kWh/m2/day in Hanoi. In the case
of direct normal irradiation the annual daily average is around
2.5 kWh/m2/day, with approximately 1,500-1,700 hours of
sunshine makes it a good site for promoting the use of PV
systems [8]. Besides, ambient temperature and wind speed are
also the main factors in PV production system.
The meteorological data embedded in PVSyst for this
project is a synthetic hourly or monthly meteorological file set
from Meteonorm in 2005 (the data collected from 1991 to
2010). It is a database containing climatological data for solar
engineering and could also be used to calculate solar radiation
on arbitrarily oriented surfaces [9, 10]. Fig. 2a shows the monthly average global irradiation of Hanoi in which the
minimum value is in January, about 67.6 kWh/m2, in contrast,
the maximum value peak at 161.6 kWh/m2 in July and August.
And the average value is about 1,391 kWh/m2 per year.
Another data source, for instance, irradiance is Climate-SAF
PVGIS which demonstrates the average daily global
irradiance (Fig. 2b) with a relatively high amount of mean
solar irradiation of 595 W/m2 in July and August and 13 hours
of sunshine.
Fig. 2. Value of average global irradiation in kWh/m2 from Meteonorm (a) and
in W/m2 from Climate-SAF PVGIS (b)
4th AUN/SEED-Net Regional Conference on Energy Engineering 2016 DRE02
43
B. Estimated energy demand and consumption
TABLE I. BUILDING ENERGY DEMAND FOR A DAY IN JUNE 2016
Load Power (kW) Use (hour) Energy (kWh)
HVAC 369.8 4 1479.2
Elevator 38.4 2 76.8
Lighting 51.6 4 206.4
Total 531.7 2193.8
Jan Feb Mar Apr May Jun
21,600 14,480 21,120 28,880 34,080 42,400
Jul Aug Sep Oct Nov Dec
36,160 37,680 37,040 24,400 31,840 31,840
TABLE III. BUILDING ENERGY CONSUMPTION IN MONTHS (KWH) 2016
Jan Feb Mar Apr May Jun
27,200 18,080 21,680 26,240 38,080 48,800
Jul Aug Sep Oct Nov Dec
44,000 40,400 --- --- --- ---
We obtained previous electric bills since 2015 as the case
study and building energy audit to determine what can be done
to reduce electricity usage. In this case, we did not take into
account the energy demand of specialized laboratory
equipment. The total amount of energy demand of the building
for this purpose was estimated about 2193.8 kWh during a day,
for example in June 2016 when energy consumption peaked, as shown in TABLE I. The energy demand consumption is
estimated based on the daily power used and its operating time.
To be certain, we studied same building's bill again by
comparing its consumption with our estimated energy demand.
As can be seen in TABLE I, II and III, building energy
consumption in one month is acceptably lower than our
estimated energy demand, thus this estimation is valid for our
purposes.
power electronic equipment such as charge-discharge
controller, inverter, and battery for energy storage and auxiliary power generation equipment. The parameter selection of every
component is very important to fulfil the system configuration.
A. PV Panels
Quantity Value
Nominal power 335 Wp
Avg. panel efficiency 17.05 %
Rated voltage 37.8 V
Rated current 8.93 A
Open-circuit voltage 47.4 V
Short-circuit current 9.62 A
Nominal operating cell temperature 46°
Power temps coefficient -0.304 %/K
Fig. 3. Effect of irradiance and temperature on power output of solar cell,
simulation results from PVSyst
temperature, solar irradiance and module type. According our
requirements, we selected crystalline silicon cells. A 15 kWp
PV system was simulated in PVSyst using the modules of XL
SW 335 rated at 335 Wp, whose specifications are shown in
TABLE IV. Based on the space availability for the correct
orientation of PV array, it was estimated that 48 modules with
total peak power 16.09 kWp that can be installed on the building’s rooftop.
To receive the maximum amount of solar radiation, all
modules are placed on the same plane, the PV array needs to
be placed at a certain angle. The placement of the PV array is
described using the plane azimuth mA and the tilt angle
m .
The tilt angle, is the angle between the horizontal plane and the
PV module. The azimuth has different definitions, but in this
paper the plane azimuth, is defined as the angle between the
orientation of the collector plane and south (in the northern
hemisphere). For geographical locations installations with
fixed modules in the northern hemisphere like Hanoi, typically
the array should be oriented to face south at 0° azimuth. The
optimal tilt angle will be presented in section V. For
simulation, two main factors of concern are incident irradiance
and cell temperature. In Fig. 3a, the maximum power of solar cell SW 335 mono-crystalline at its NOCT (45°C) is
proportional to the incident irradiance. One panel can generate
306.7 W under incident irradiation 1,000 W/m2. The generated
electric power drops quickly with decreasing incident
irradiance, when the irradiance is 200 W/m2, the panel can
only produce 59.6 W. In Fig. 3b, the power output depends
strongly on temperature with negative coefficient. A PV cell
generated power from 269 W to 356 W under 1,000 W/m2
incident irradiance when the cell operating temperature varies
from 10oC to 70oC.
B. Inverter and matching PV array/Inverter
The inverter takes the DC power from the PV array and
converts it into standard AC power used by the building
appliances. Some factors that we have to take into account in
choosing inverter include:
- The generated power of PV array: The power of inverter
is the output power of system. When the power of inverter is
higher than generated power from the array, it increases cost
unnecessarily. On the contrary, when the array generates more
than the power that inverter can handle, the exceed power is
waste and the efficiency of the system decreases.
4th AUN/SEED-Net Regional Conference on Energy Engineering 2016 DRE02
44
- Voltage and current ranges: The highest output voltage
and current of the array have to be under the peak voltage and
current that inverter can handle. In operation mode, basing on
the highest and lowest produced Vmpp of the array, we have to
select inverter that having a suitable operating range. The DC
input current range of inverter has to be considered similarly to
be suitable with the DC output current of array.
- Energy appliances and installation: There are two kinds of
inverter: Central inverter and string inverter. Central inverter is suitable for a homogenous system when all panels have the
same orientation and the same DC output voltage. This kind of
inverter can reduce the cost per unit of power due to the better
performance and footprint size (in comparison with string
inverter). However, when strings work in different conditions,
the mismatch among them will reduce the efficiency of
harvesting system. The complication of high DC voltage
wiring could be another problem, which increases the risk of
fire. In the other hands, string inverter is suitable for the system
that contains various modules with different orientations.
The simulation only consider about choosing which type
of PV module and inverters. PVSyst has a built in system that
matches the number of inverters with the number of strings. It
also proposes a number of modules in each series, and the number of strings based on the sizing parameter. This is to
assure that all requirements regarding current, voltage and
power levels are met. When an inverter have multiple MPPT
inputs, in PVSyst, it divide the operating power between the
inputs. Two inputs take half the nominal power at each input.
It is not possible to share the power unequally between each
MPPT input. The main reason for choosing two different
modules of Sunmodule XL SW 335 and two kind of inverters
is to compare the performance of different types of well
known (TABLE V). These technique conditions are used
together with the numbers proposed by PVSyst to set design
the configurations in (TABLE VI). The ratio between peak power of array (in standard testing
condition 1,000 W/m2 irradiance and 25°C) and the power of
inverter (15 kWp) is 1.07. This ratio is acceptable since the
high efficiency of inverter choosing, and the high insolation in
Hanoi [11, 12].
Inverter Number Nominal
50A (2 inputs
TABLE VI. SIMULATED ARRAY/INVERTER COMBINATIONS
Inverter Num. Mod. in
series String MPPT input/String
Several studies have suggested that battery storage co-
located with grid-integrated building-scale PV generation
benefits electricity distributors in maintaining system voltages
within acceptable limits [13]. Energy storage can supply more
flexibility and balancing to the grid, providing a back-up to
intermittent solar energy. Locally, it can improve the
management of distribution network, reducing costs and
improving efficiency. The design principle for sizing the
system’s battery is to compensate for daily variations in solar
radiation and not to act as seasonal energy storage [14]. PV
modules are not an ideal source for battery charging. The PV output is heavily dependent on weather conditions, therefore an
optimum charge/discharge cycle can be guaranteed, resulting in
a low battery state of charge (SOC) [15]. Low battery SOC
drives to shorten the life of the battery. The battery proposed is
Lithium battery LiFePO4 with 100 Ah battery capacity, with
three blocks of 5 kWh/each. Then the nominal voltage for each
cell is 54 V. The battery has columbic efficiency by 97%.
V. PV SYSTEM PERFORMANCE ANALYSIS
The performance evaluation of this study is based on the
PVSyst software simulation. This part discuss the output from
the simulation tool such as the potential energy resources, the
optimum tilted angle of PV array and energy production from
the PV array as well as system losses.
A. Tilted angle optimization of PV array installation
To optimize the tilted angle m for the rooftop PV array, an
ideal case was built with azimuth equal to 00mA . The tilted
angle of the free standing system was adjusted from 0° to 30°.
The result of simulation is shown in Fig. 4 and 5. The specific
energy product and system energy product indicates that the
best tilted angle is 15°, however the difference between the
results is small for the range of tilt angles of 9, 12, 15, 18, and
21 degree. Therefore, the tilt angle of South faced PV system
for this building can be flexible between 9 and 21 degree. In
reality, optimum PV array output can also be achieved using
tilt angle approximately equal to the site’s latitude of 21°N for
Hanoi. Mono crystalline cell is better choice than poly
crystalline cell. Central inverter is better choice than string inverter. With the mono-crystalline and central inverter,
simulation shows an annual system production of PV 15 kWp
system of 19,348 kWh/year, and the specific yearly production
is 1,205 kWh/kWp/year. The overall performance ratio (PR) is
0.849 and the normalized production is 3.3 kWh/kWp/day.
The PR is a measure for the overall losses of a PV system and
is defined as the ratio of final energy yield of the PV system in
kWh/kWp to a reference yield, which takes only solar
irradiation into account.
office equipment like computers, laptops and three phase
voltage to supply the elevators and HVAC system, therefore three separable single phase inverters configuration is chosen.
The following section presents the analysis of simulation
results to evaluate the system performance for the mono-
crystalline and string inverter variant.
4th AUN/SEED-Net Regional Conference on Energy Engineering 2016 DRE02
45
Fig. 4. Specific product of PV system for 00mA by tilted angles
Fig. 5. System product of PV system for 00mA by tilted angles
B. Analysis of power output and energy production
Using an optimally orientated PV array with 00mA and
015m , PV mono-crystalline modules and string inverter are
mentioned in this simulation, when the yearly data is averaged
over 24 h period for each month by estimated in PVSyst, we
get the average power production of PV array profile for
different months as shown in Fig. 6. It can be observed that:
- The average monthly peak power ranges between 8.7 kW
in August and 5.2 kW in January. This indicates that the PV
array on an average only produces around 60 % of its rated power even in the sunniest month of the year. The rated power
of the PV array using 48 modules of XL SW 335 is estimated
under standard test conditions, i.e. 1,000 W/m2 and 25°C.
However, according to Fig. 2, the average global irradiance in
August is 595 W/m2 and the ambient temperature is about
37.5°C.
- PV array generates energy for 13 hours during summer
months while it is only 11 hours in winter.
Fig. 6. Average power output of 15 kWp PV array as a function of time of the
day for different months.
Fig. 7. Average daily yield of 15 kWp PV array for different months.
Fig. 7 and 8 show the average daily yield of PV system for
different months. The different of PV yield of two time can be
observed between January and August. In August, PV yield
always reaches a high level with almost no fluctuation around
the average value approximately 64.290 kWh/day. There are
also cloudy days with low daily yield about of 30 kWh which
is showed in these months from May to November. By
contrast, in January, PV yield fluctuates dramatically among
days in month. There are cloudy days with low daily yield of
< 5 kWh and sunny days with yield > 70 kWh. This trend also occurs in the winter: in December, January, February and
March.
Fig. 8. Average daily yield of 15 kWp PV array for January and August.
4th AUN/SEED-Net Regional Conference on Energy Engineering 2016 DRE02
46
PV array failed to generate 100% the energy delivered
from the sun because of some losses. Fig. 9 explored the
overall system loss diagram for site installation. A collector
plane received about 1,391 kWh/m2 of global incident
irradiation. But the effectiveness plane receives the irradiance
only at 1,366 kWh/m2. The biggest losses happen in PV array
production energy. Energy produced from the PV array
affected by several factors such as ambient temperature which
takes the main part of loss (9.4%), solar incidence, manufacture
mismatch and ohmic wiring. The loss in inverter is 3.1% since
the efficiency of inverter in specification of producer is 96.9%. Other losses in inverter operation is zero, shows that PV array
and inverter perfectly fit each other. The actual energy supplied
to the load can identify after through the losses of regulator and
battery storage. These annual system production of PV 15
kWp system of 19,138 kWh/year.
Fig. 9. Overall system losses in case: azimuth at 0° and tilted angle at 15°,
simulation result from PVSyst [9, 10]
VI. CONCLUSION
This paper presents the use of PVSyst as a simulation tool
to evaluate the PV array installed on building’s rooftop. This
software enables the PV project designer to select system
configurations as well as predict its energy production. System
sizing depends strongly on the geographical location of the
installation site. This work studied a PV system which is
applied for low-voltage grid connection at building-scale, with electrical storage system in the scope of our research on energy
efficiency dedicated to the green building. The simulations
shown that the average monthly peak power ranges from 8.7
kW in August to 5.2 kW in January. This indicates the PV
array on an average only produces of 60 % of its rated power
even in the sunniest month of the year. It is also helpful as a
pre-decision of installation. The results indicated that the
optimal tilt for PV modules installed on building’s rooftop in
Cau Giay-Hanoi to get maximum yield is 15°.
For future works, the sizing of grid-connected PV system
should be extended to all component and an economic analysis
should also be performed. In addition, measurement equipment
such as pyrometer can be used to obtain on-site irradiance data
in real time. Such measuring equipment could be used monthly in order to correct the simulation results provided by PVsyst
with the output and real production from the PV installation.
REFERENCES
[1] L. Pérez-Lombard, J. Ortiz and C. Pout, “A review on buildings
energy consumption information,” ELSEVIER, March 2007.
[2] S. Deng, R. Z. Wang, Y. J. Dai, “How to evaluate performance of net zero energy building e A literature research,” Energy, vol. 71, 2014, pp. 1-16.
[3] A. Scognamiglioa, G. Adinolfia, G. Graditia, E. Sarettab, “Photovoltaics in net zero energy buildings and clusters: Enabling the smart city operation,” Energy Procedia 2014, vol. 61, pp. 1171–1174.
[4] C. N. Chong, E. Sidgwick, A. Wanniachchi, D.-T. Bui, H. Lee, A. Tuan, “Viet Nam: Energy sector assessment, strategy, and road map,” Asian Development Bank report, December 2015.
[5] D. L. Nguyen, “A critical review on Energy Efficiency and Conservation policies and programs in Vietnam,” Renewable and Sustainable Energy Reviews, vol. 52, 2015, pp. 623–634.
[6] H. T. Nguyen, M. Gray, “A review on Green Building in Vietnam,” Procedia Engineering, vol. 142, 2016, pp. 314-321.
[7] Y. Millet, “How to promote Green Building in Vietnam?,” Green Housing Conference- Goethe institute, October, 2012.
[8] J. Polo, A. Bernardos, A. A. Navarro, C. M. Fernandez- Peruchena, L. Ramírez, María V. Guisado and S. Martínez, “Solar resources and power potential mapping in Vietnam using satellite-derived and GIS-based information,” Energy Conversion and Management, vol. 98, 1 July 2015, pp. 348–358.
[9] A. Mermoud, “Use and Validation of PVSYST, a User-Friendly Software for PV-System Design,” European conference; 13th, Photovoltaic solar energy; 1995; Nice; France.
[10] S. Muller, J. Remund, S. Kunz, and C. Schilter. Handbook part 1: Software. Meteonorm, version 6 edition, December 2012.
[11] J. D. Mondol, Y. G. Yohanis, B. Norton, “Optimizing the economic viability of grid-connected photovoltaic systems,” Applied Energy, vol. 86, 2009, pp. 985–999.
[12] G. Notton, V. Lazarov, L. Stoyanov, “Optimal sizing of a grid- connected PV system for various PV module technologies and inclinations, inverter efficiency characteristics and locations,” Renewable Energy, vol. 35, 2010, pp. 541–554.
[13] E. L. Ratnam, S. R. Weller, C. M. Kellett, “An optimization- based approach to scheduling residential battery storage with solar PV: Assessing customer benefit,” Renewable Energy, vol. 75, 2015, pp. 123-134.
[14] E. Galvin, P. K. W. Chan, S. Amstrong and W. G. Hurley, “A stand-alone photovoltaic supercapacitor battery hybrid energy storage system,” 13th International Power Electronics and Motion Control Conference (EPE-PEMC 2008).